18 ON GYMNOSPEHMATODS FRUITS 
lateral canals retain their original dimensions ; and the ripe seed 
shows the unicellular cavity with its single seed and relatively slender 
lateral canals. An examination of the early stage of the scale would 
very readily give the erroneous impression that it contained two seeds, 
or, rather, that the hard integument was bilocular, and contained a 
single seed in each loculament. 
Besides the species lately introduced by nurserymen which have not 
vet been satisfactorily described, the genus Araucaria contains six 
well-known species, four of which are natives of Polynesia, and the 
remaining two of South America. The species thus geographically 
grouped have so many peculiarities in common, that Salisbury pro- 
posed to establish two genera for them, — the one, Cotymbeia, for the 
two American species, characterized, as regards the cone, by having 
the scales without wings; and the other, Eutassa (Eutacta, Link), for 
the Polynesian species, which have wings to the scale. Although 
these and the other characters obtained from the number of the 
anthers and cotyledons, the form of the leaves, and the germination, 
are of importance, they have not been considered by recent systematists 
of generic value, but sufficient only for the division of the genus into 
two natural sections. Another peculiarity is possessed by the majo- 
rity of the Polynesian species, which is important in connection with 
our fossils. The Brazilian species never exhibit externally any division 
of the scale into an upper and under portion. The small upper scale, 
so evident in A. Bidwilli (Plate LX. Pig. 4) and A. exceha, and in 
the two fossils, is so reduced in them that it is only discoverable on 
the upper surface of the scale after it has been withdrawn from the 
cone.* This small upper scale is larger in the fossil than in any of 
the recent species. Three different views are entertained regarding 
the nature of this portion of the scale. Bichardt and EndlicherJ 
describe it as an appendage to the seed; the late Sir W. J. Hooker§ 
supposes it to be the dilated "upper base" of the scale-leaf folded 
down upon its upper surface; and Prof. Dickson || holding that the 
mnaris 
from New Caledonia. The scale of this species has perhaps the most largely- 
developed wings, but the small upper scale is even more reduced than in the 
American species. 
t Richard, * Memoires sur les Coniferes/ p. 87. 
X Endlieher, ' Synopsis Coniferarum,' p. 184. 
§ London Journ. of Botany, vol. ii. p. 504. 
II Edinburgh New Phil. Journ., 1861, p. 197. 
